Date of Award

Summer 6-30-2015

Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Sally Wallace

Second Advisor

Yongsheng Xu

Third Advisor

Shiferaw Grumu

Fourth Advisor

Abdu Muwonge


The three essays of this dissertation inform tax policy design. It is a compilation of empirical and experimental research work. The first and the second essays explore the performance of a recent tax policy reform at the sub-national level in India in terms of revenue efficiency as well as economic efficiency. India is among the only three countries in the world to have adopted a sub-national VAT. Therefore, empirically examining its performance not only improves the understanding of this important tax policy reform but also informs tax policy decision-making at the sub-national level in other developing countries.

India transitioned to the state-level VAT between the years 2003 and 2008. Among other things, it was expected to achieve revenue growth and decrease tax cascading on commodities by improving economic efficiency of the indirect tax system. In the first essay, I model the impact of the VAT on revenue by adding revenue dependent administrative and compliance costs associated with taxation to an existing model developed by Keen and Lockwood (2010). The theoretical results show that replacing one type of indirect tax with another improves long-run revenue efficiency only if there is a net decrease in the administrative, compliance and distortionary costs of taxation at the margin. I then compile a unique state-level dataset for the years 1990 to 2010 to determine changes in the long-run revenue efficiency from the use of the VAT. This essay contributes to the literature by extending an existing revenue efficiency model and testing it in the unique situation of India’s sub-national VAT. The results reveal a significant improvement in the long-run revenue efficiency of the sales tax instrument used by state governments. The model implies this improvement is driven by a net fall in the marginal taxation costs from the use of the state-level VAT. This finding has important implications on the role of a sub-national VAT in the future as an effective tax instrument in the developing countries.

The second essay appeals to the general theory of tax incidence which suggests that a VAT will have less impact on prices than a traditional turnover tax because the VAT does not “get stuck” in the production process as a turnover tax does. The impact should be larger for goods that have more components to the production process as the tax then “touches” more of the final product. In this essay I measure the change in the level of tax cascading with VAT by using multiple waves of the state- and household-level expenditure surveys. Specifically I test the impact of the VAT on the real consumption of households on a variety of consumption goods. I find the biggest significant decrease in the tax cascading burden of the long-term durable goods which essentially involve the maximum production components. This result is found in the 18 more developed states of India which are the focus of the empirical analysis due to data constraints.

The third essay is an experimental research which looks at the influence of institutions on the economic burden of an excise tax. The traditional long-run tax incidence theory establishes that the economic incidence of an excise tax is independent of the assignment of the liability to pay tax. However, the theory is silent on the possible effects of the market institutions on tax incidence. Since all markets need an institution to function and every market institution has its own unique price and quantity determination property, it is important to understand its bearings on the incidence of taxes. Existing experimental research has tested economic incidence under many different market institutions but no previous research systematically analyzes and compares the incidence of a unit tax under two important market institutions we deal with in everyday life. One of these institutions is posted offer which dominates the consumer goods markets in developed countries and the other is double auction which is frequently observed in developing countries. I report a significant impact of these market institutions on tax incidence. In particular, I find that consumers bear a much higher burden of a unit tax in the posted offer markets as compared to the double auction markets and their burden further increases when the liability to pay the tax is on the seller.